AN EVALUATION OF THE RELATIVE EFFICACY OF … manner in which teachers mediate children’s learning varies across ... direct instruction, discovery learning, ... its reliance on contrived

AN EVALUATION OF THE RELATIVE EFFICACY OF AND CHILDREN’SPREFERENCES FOR TEACHING STRATEGIES THAT DIFFER IN

AMOUNT OF TEACHER DIRECTEDNESS

NICOLE A. HEAL, GREGORY P. HANLEY, AND STACY A. LAYER

UNIVERSITY OF KANSAS

The manner in which teachers mediate children’s learning varies across early childhoodclassrooms. In this study, we used a multielement design to evaluate the efficacy of threecommonly implemented strategies that varied in teacher directedness for teaching color- andobject-name relations. Strategy 1 consisted of brief exposure to the target relations followed by anexclusively child-led play period in which correct responses were praised. Strategy 2 was similarexcept that teachers prompted the children to vocalize relations and corrected errors via modelprompts. Strategy 3 incorporated the same procedures as Strategy 2 except that a brief period ofteacher-initiated trials was arranged; these trials involved the use of prompt delay betweenquestions and prompts, and correct responses resulted in tokens and back-up activity reinforcers.Children’s preferences for the different teaching strategies were also directly assessed. Strategy 3was most effective in promoting the acquisition and generalization of the color- and object-namerelations and was also most preferred by the majority of children, Strategy 1 was the leasteffective, and Strategy 2 was typically the least preferred. Implications for the design of earlyeducational environments based on evidence-based values are discussed.

DESCRIPTORS: child preference, direct instruction, discovery learning, embeddedteaching, evidence-based teaching strategies

_______________________________________________________________________________

The practices of early childhood educatorshave been guided by the recommendations fromboth the National Association for the Educationof Young Children (NAEYC; Bredekamp &Copple, 1997) and the Division of EarlyChildhood (DEC; Smith et al., 2002). Adominant belief that is consistent with theseposition statements is the importance of variedchild–environment interactions to promotelearning. Wolery and Wilbers (1994) outlineda continuum on which teaching strategies canbe located, with exclusively child-initiatedinteractions at one endpoint and exclusivelyteacher-initiated interactions at the other end-point. Strategies located at the child-initiated

end of the continuum generally result in highlevels of child engagement, whereas strategieslocated at the teacher-initiated end generallyproduce specific teacher-selected behaviors(Wolery & Sainato, 1996). Discovery learning,embedded teaching, and direct instruction arethree specific early childhood teaching strategiesthat occupy different points on this continuum.

On the child-initiated endpoint of thecontinuum lies discovery learning, a teachingmethod developed from the constructivistphilosophy of learning (Piaget, 1970) in whichthe learner is expected to discover new ideas andrelations through independent interactions withthe environment with little or no guidance froma teacher (Bruner, 1961). The teacher’s primaryrole is to arrange the environment to promoteindependent interactions with the materials andexpose the child to the learning objectivesthrough intermittent commenting and ac-knowledgment when a child is successful (Klahr& Nigam, 2004; Solter & Mayer, 1978). Activeengagement and reinforcement that are notsocially mediated (i.e., so-called intrinsic moti-

This research is based on a dissertation by the firstauthor submitted to the University of Kansas in partialfulfillment of the requirements for the PhD degree. Wethank Edward Morris, Jan Sheldon, Holly Storkel, andRachel Thompson for their helpful comments on anearlier version of this manuscript.

Requests for reprints should be addressed to Gregory P.Hanley, Psychology Department, Western New EnglandCollege, 1215 Wilbraham Road, Springfield, Massachu-setts, 01119 (e-mail: [email protected]).

doi: 10.1901/jaba.2009.42-123

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2009, 42, 123–143 NUMBER 1 (SPRING 2009)

123

vation) are presumed to be the critical variablesunderpinning the learning that occurs in thisapproach.

Embedded teaching (Bricker, Pretti-Front-czak, & McComas, 1998) is located in themiddle of the continuum and is characterizedby instructions and feedback regarding targetskills being delivered within child-initiatedactivities during typical routines. Embeddedteaching strategies were derived from the earlywork of Hart and Risley (1968, 1975) onincidental teaching. A typical example ofincidental teaching involves the teacher placingpreferred materials within sight but out of thechild’s reach, thereby increasing the likelihoodthat the child will request the materials. Whenthe child engages in the desirable request, thematerials are provided. According to Daugh-erty, Grisham-Brown, and Hemmeter (2001),multiple variations of the embedded teachingprocedure have been described, with character-istics such as type of activity, prompts,programmed consequences, and learning mate-rials distinguishing the variations. Nevertheless,learning opportunities are considered childinitiated, and the reinforcing consequences forengaging in target responses are considerednatural in that the child continues to play withmaterials following a trial of embedded teach-ing, or the child receives the requested item thatis associated with the scheduled activity.

A third teaching strategy, which is located onthe teacher-initiated end of the continuum, isdirect instruction, in which the teacher plays amore prominent role in the teaching situation(Magliaro, Lockee, & Burton, 2005). Directinstruction is characterized by relatively simpleand precise materials tailored to specificlearning objectives, planned (and sometimesscripted) prompting procedures, provision ofhigh-quality reinforcers for correct responding,and multiple trials conducted during briefteaching periods (Fredrick, Deitz, Bryceland,& Hummel, 2001). Wolery and Sainato (1996)outlined a variety of procedures that are often

adopted during direct instruction; these includeconstant or progressive prompt delays, errorcorrection via modeling correct answers or brieftime-out periods, and differential reinforcementof correct responding with high-quality orhighly preferred items.

Each of these three teaching approachesvaries primarily in the amount of teacherdirectedness during the teaching situation, andeach has unique strengths and weaknesses. Thediscovery-oriented approaches require the care-ful selection of learning materials, but demandson the teacher are not high during the learningperiod. A second advantage of this approach isthat the teacher respects the momentarypreferences of children for simultaneouslyavailable activities and materials. However,because this approach is devoid of promptsand feedback from the teacher, specific learningobjectives are difficult to target, and it isdifficult to determine the specific skills acquiredas a direct function of this teaching approach(Mayer, 2004). Despite the lack of empiricalevidence supporting the efficacy of purelydiscovery-oriented teaching methods (Mayer),these methods are included as recommendedpractice by the NAEYC (Bredekamp & Copple,1997).

Because teacher prompting and feedback arearranged during embedded teaching, specificskill acquisition occurs with this approach (Fox& Hanline, 1993; Horn, Lieber, Li, Sandall, &Schwartz, 2000; Woods, Kashinath, & Gold-stein, 2004). When compared to direct instruc-tion, embedded teaching produced similar skillacquisition; however, better skill generalizationwas observed by Losardo and Bricker (1994)and McGee, Krantz, and McClannahan (1985)with the embedded procedures. A primarydisadvantage of embedded teaching is thedifficulty inherent in routine and successfulimplementation. For example, Pretti-Frontczakand Bricker (2001) observed that followingextensive training on implementing embeddedteaching, early childhood and early childhood

124 NICOLE A. HEAL et al.

special educators implemented these procedureson less than 10% of observation intervals.

Along with questionable generalization ofskills acquired during direct instruction, aprimary criticism is usually directed towardsits reliance on contrived learning opportunities,materials (e.g., flash cards), and programmedconsequences (e.g., tokens, traded in for backupmaterial reinforcers; see Strain et al., 1992).Despite strong empirical evidence for theefficacy of direct instruction (Adams & Engle-mann, 1996; Stebbins, St. Pierre, Proper,Anderson, & Cerva, 1977), these proceduresare not widely adopted by early childhoodeducators (Walsh & Petty, 2006).

A teaching strategy that has overwhelmingempirical support but that is not considered tobe socially acceptable by relevant consumers(parents, teachers, and interventionists) is notlikely to be adopted in practice (Wolf, 1978).Schwartz (1999) observed the importance of theconstruct of social validity with the earlychildhood interventions of direct instructionand activity-based (embedded) instruction forteaching arithmetic. The strongest researchsupport is in favor of direct instruction, butembedded teaching has greater social validity(i.e., parents and teachers find it more accept-able), and the latter is adopted more inpreschool settings. Thus, determining theacceptability of early childhood practices, whichis usually accomplished by administering ques-tionnaires to teachers or other relevant stake-holders, is important when designing earlychildhood classroom practices. However, anadditional measure of a practice’s value may beobtained from the children who directlyexperience it, and these measures of children’sacceptance may provide additional compellingevidence for the adoptability of a practice.

Determining the acceptability of instruction-al strategies with young children of limitedverbal competence, limited history with thestrategies in question, or both, complicates thisprocess considerably. Nevertheless, a procedure

for directly determining preferences of childrenwith disabilities for behavioral interventions wasdescribed by Hanley, Piazza, Fisher, Contrucci,and Maglieri (1997) and Hanley, Piazza, Fisher,and Maglieri (2005). These procedures wererecently extended to determine children’spreferences for instructional contexts that variedin the amount of child control (Tiger, Hanley,& Hernandez, 2006), amount of informationregarding the availability of teacher attention(Tiger, Hanley, & Heal, 2006), and by the typeof motivational system (Heal & Hanley, 2007).In these studies, different-colored poster boardswere correlated with the different teachingstrategies (or interventions), and children re-peatedly experienced the strategies in thepresence of the colored poster boards. Smallercolored cards or microswitches, one associatedwith each strategy, were then made available tothe children outside the room in which teachingtypically occurred, and the child was asked toselect the one he or she liked best. When thechild handed a card to the teacher (or pressed amicroswitch), the teacher and child entered theroom and briefly experienced the strategyassociated with the selected color. This processof handing cards (or pressing switches) andexperiencing correlated strategies was repeateduntil the child selected one option on a regularbasis (or some other pattern emerged). Thus,preferences for contexts, which were difficult todescribe to young children, were directlyassessed by recording each child’s selections ofcues correlated with the important teaching orintervention strategies.

In the current study, the relative efficacy ofand children’s preferences for three teachingstrategies that differed in the amount of teacherdirectedness (discovery oriented, embedded,and direct instruction) was determined usingsingle-subject experimental designs. The proce-dures implemented in each teaching strategywere not mutually exclusive; instead, moreelements were added to each successive strategysuch that the mediating role of the teacher in

EFFICACY AND PREFERENCE 125

the learning situation was more extensive acrossthe strategies. Thus, our analysis allows adescription of the added value of embeddinglearning opportunities and providing briefperiods of direct instruction. Relative efficacyof these strategies for teaching preschoolchildren naming relations was assessed byexamining skill acquisition data (i.e., numberof learning opportunities, percentage of correctresponses, and latency to mastery) and pre- andposttest data with respect to the relations taughtin each strategy. In addition, children’s prefer-ences for the strategies were determined bydirectly measuring their selections of eachstrategy over time.

METHOD

PARTICIPANTS AND SETTING

Participants were 6 Caucasian English-speak-ing children, 4 girls and 2 boys aged from 48 to61 months (M 5 55 months). The childrenattended a full-day inclusive preschool class-room that served children of typical andatypical development. None of the childrenhad been diagnosed with a developmentaldisability; however, individualized curriculashowed great variability in their progress.Children were selected for participation basedon informed consent and consistent classroomattendance. All sessions were conducted in asmall room (3 m by 3 m) near the children’sclassroom that contained a child-sized table andchairs (in addition to the session materials).

MATERIALS

Two relations were taught to each child. Thechild was initially taught to vocally label colorsin Spanish; then he or she was taught to vocallylabel animals in Spanish. There were 12 colorand 12 animal names taught. Each strategy wasassociated with four color relations and thenfour animal relations (see Table 1 for thespecific color and animal relations). Three setsof materials designed to evoke the targetresponses (color or animal names) during each

assessment (i.e., color- and animal-name assess-ments) were rotated across sessions. Withineach set of materials, each target stimulus wasrepresented by three items. We included threedistinct sets of materials for each relationbecause we aimed to keep the children’s interestin the toy sets high throughout the study, wedid not want the children to select one of thethree teaching strategies to gain access to aparticular toy set during the preference assess-ments, and we were explicitly programming forgeneralization of the name relations acrossstimuli. In addition to the multiple toy sets,12 color and animal cards, plastic tokens, and atreasure box were arranged in the strategyinvolving some direct instruction. Each teach-ing strategy was associated with distinctlycolored large (60 cm by 75 cm) and small(15 cm by 10 cm) laminated poster boards.

RESPONSE MEASUREMENT AND

INTEROBSERVER AGREEMENT

Data were collected using paper and pencilwithin 15-s intervals. A child-initiated learningopportunity was defined as the first occurrencewithin each 15-s interval of the child grasping atarget item or pushing down on the target itemfor a minimum of 1 s. If a second target item

Table 1

Target Responses for Each Teaching Strategy

Relation

Target responses

Strategy 1 Strategy 2 Strategy 3

Color Anaranjado Morado AmarilloOrange Purple Yellow

Azul Negro BlancoBlue Black WhiteGris Plata CafeGray Silver BrownRosa Rojo VerdePink Red Green

Animal Caballo Pato CochinoHorse Duck Pig

Gallina Perro PajaroChicken Dog Bird

Gato Serpiente RanaCat Snake FrogPez Tortuga VacaFish Turtle Cow


was touched within the same 15-s interval, thiswas not scored as a child-initiated learningopportunity. Teacher-initiated learning oppor-tunities were scored when the teacher held up acolor or animal card and in English said, ‘‘Whatcolor [animal] is this?’’ Learning opportunitiesare reported as a frequency count within eachteaching strategy, and the numbers of child-initiated and teacher-initiated learning oppor-tunities are combined in Strategy 3. Given anoccurrence of a learning opportunity, a correctresponse was scored when the child indepen-dently and correctly said the Spanish wordcorresponding with the target stimulus within5 s of the initiation of the learning opportunity,and a frequency count across intervals isreported. In addition, the mean percentage ofcorrect responding was calculated by calculatingthe mean percentage correct scored within thefinal five sessions of each teaching strategy forthe color- and animal-name assessments.

During the preference assessments, selectionswere scored and defined as the child removingone of the three cards from the door andhanding it to the teacher. Card selections arereported as a preference rank in which 1represents the first card selected and 3 repre-sents the last card selected. Session duration wasrecorded and is reported as mean durationacross children and assessments. Mean sessionduration was calculated by calculating meansession durations for each child; individualmeans were then averaged across all childrenand assessments.

A second observer recorded behavior simul-taneously but independently in at least 27% ofsessions across all children and assessments(range, 27% to 60%). The records of eachobserver were compared on an interval-by-interval basis. An agreement was scored whenboth observers scored the occurrence or nonoc-currence of a learning opportunity and occur-rence or nonoccurrence of a correct responsewithin each interval. Interobserver agreementwas calculated by dividing the number of

agreements by the number of agreements plusthe number of disagreements and then convert-ing this ratio to a percentage. Agreement was94% (session range, 50% to 100%) for learningopportunities and 98% (session range, 67% to100%) for correct responses across all childrenand assessments. Interobserver agreement wascollected and calculated in the same manner asdescribed above for card selections in aminimum of 20% of all preference assessmentsacross all children (range, 20% to 37%). Anagreement, defined as both observers recordingthe same card selection for each session, was100% across all children and assessments.

FIDELITY OF TEACHERS’ PROMPTS

AND CONSEQUENCES

Measures of procedural fidelity were collectedon the teachers’ delivery of the initial vocalprompt and consequences provided followingthe child’s response in each teaching strategy.Given a learning opportunity, data wererecorded on the nonoccurrence (Strategy 1) oroccurrence (Strategies 2 and 3) of an initialvocal teacher prompt (e.g., ‘‘What color isthat?’’) for each session. Because the teacher wasnot to deliver a vocal prompt following theinitiation of a learning opportunity duringStrategy 1 sessions, to calculate fidelity thenumber of learning opportunities in which theteacher did not deliver a vocal prompt wasdivided by the total number of learningopportunities, and this ratio was converted toa percentage. By contrast, the teacher wasrequired to deliver a vocal prompt duringStrategies 2 and 3; thus, to calculate fidelity,the number of learning opportunities in whichthe teacher delivered a vocal prompt wasdivided by the total number of learningopportunities, and this ratio was converted toa percentage.

Given a learning opportunity and childresponse, consequences also varied across thethree teaching strategies. Following an incorrector no response the teacher was required not to


deliver a model prompt during Strategy 1sessions, but the teacher was required to delivera model prompt following an incorrect or noresponse during Strategies 2 and 3. In addition,the teacher was always required to deliver praisefollowing a correct response in all threestrategies. To calculate fidelity measures onteacher consequences during Strategy 1 given alearning opportunity, the number of correctresponses in which the teacher delivered praiseplus the number of incorrect or no responses inwhich the teacher did not deliver a modelprompt were divided by the total number ofchild responses and then converted to apercentage. For Strategies 2 and 3, given alearning opportunity, the number of correctresponses that were followed by teacher praiseplus the number of incorrect or no responsesthat were followed by a teacher model promptwere divided by the total number of childresponses, and this ratio was converted to apercentage.

Across all children and assessments, theteacher did not deliver a vocal prompt followinga learning opportunity for a mean of 99% ofopportunities during Strategy 1. The teacherdid deliver a vocal prompt following theinitiation of a learning opportunity on a meanof 93% and 97% of opportunities duringStrategies 2 and 3, respectively, across allchildren and assessments. Given a learningopportunity and child response, the teacherdelivered the appropriate consequence on amean of 99%, 96%, and 98% of opportunitiesacross all children and assessments duringStrategies 1, 2, and 3, respectively. Takentogether, these data suggest that the proceduresof each strategy were implemented with a highdegree of fidelity.

EXPERIMENTAL DESIGN

A multielement single-subject experimentaldesign (Sidman, 1960) was used to determinethe relative efficacy of the three strategies forteaching color- and object-name relations to 6

preschoolers. The counterbalanced and rapidalternation of the three teaching conditionsallowed performance in each of the threestrategies to be influenced by outside factors(lack of sleep, illness) similarly and for eachchild to experience each strategy for the sameamount of time. A concurrent-chains arrange-ment (Catania & Sagvolden, 1980; Hanley etal., 1997) was used to determine children’spreferences for the teaching strategies.

PROCEDURE

Overview

Three preassessments were conducted priorto evaluating the relative efficacy of andpreference for the three teaching strategies. Toidentify the colors that would be associated witheach strategy, a paired-item color preferenceassessment was conducted first. To ensure thateach child had the necessary skills to echo amodel prompt, an echoic assessment wasconducted second. To assess each child’s skilllevel with respect to the color- and animal-namerelations, a pretest with all 12 target relationswas conducted third. Following the simulta-neous evaluation of efficacy and preference, posttests of the color- and animal-name relationswere conducted and served as a measure ofgeneralization for the acquired relations.

Preassessments

Color preference assessment. Ten colored cardswere initially included in a paired-item assess-ment (Fisher et al., 1992) with each child. Eachcolored card was paired with every other coloredcard once, the pairs were presented to the childone at a time, the child was prompted to touchthe color he or she liked best, and the coloredcard selected was scored. The order of presentedpairs was randomized. For every card selection,the teacher delivered a brief statement of praise(e.g., ‘‘thanks’’); therefore, no differentialconsequences were provided for selecting aparticular colored card. Selection percentagesand a preference hierarchy were obtained by


dividing the number of times a card wasselected by the total number of times it waspresented, and this ratio was converted to apercentage. The three colors that were identifiedas moderately and similarly preferred (i.e.,colors that were identified in the middle ofthe preference hierarchy) were selected todecrease the likelihood that the child’s selectionsin the strategy preference assessments might becontrolled by a preexisting color preference.The three cards were then randomly assigned toeach of the strategies and were held constantthroughout the efficacy and preference assess-ments for each child.

Echoic assessment. A nine-trial assessment todetermine if each child could echo one- to five-syllable words in English was conducted next.The trial types were adjusted according to theresponses. In the first trial, the teacher vocallymodeled a one-syllable word (e.g., ‘‘cat’’ or‘‘milk’’) and then prompted the child to repeatthe word. If the child successfully echoed theword, a two-syllable word (e.g., ‘‘apple’’ or‘‘crayon’’) was presented on the next trial. Thisprocess continued until five-syllable words werepresented. If a response was incorrect or thechild did not respond within 5 s of the modelprompt, the next word presented contained oneless syllable than the word that was notsuccessfully echoed. If all responses werecorrect, the teacher vocally modeled five-syllablewords for the final four trials. A short statementof praise was delivered following correctresponses, and no corrective feedback wasprovided following incorrect or no responses.The next trial was initiated following a 2-spause. All children correctly echoed at least onefive-syllable word during this assessment.

Pretest. A novel teacher who was fluent inSpanish conducted all pretests with each child.The teacher instructed the child to answer all ofthe questions in Spanish prior to the start ofeach pretest. Sitting across from the child at achild-sized table, the teacher held up onelaminated color or animal card and in English

asked, ‘‘What color [animal] is this?’’ Noconsequences were delivered following corrector incorrect responses. However, the teacherprovided statements of descriptive praise (e.g.,‘‘nice sitting’’) after every other trial. Correctand incorrect responses were scored.

Efficacy Assessment

After the three preassessments were conducted,the efficacy assessment began. Three distinctsessions made up a session block; one sessionblock was conducted per day, such that each childexperienced each teaching strategy once daily.The same teacher conducted all sessions andprovided some form of attention (i.e., specificprompts, praise, comments) during each 15-sinterval such that the amount of attention thechild received was similar across the threestrategies. The mediating role of the teachervaried across the three strategies, ranging fromplaying a very minimal role (Strategy 1) to playinga more prominent role (Strategy 3) in the teachingsituation. In addition, the strategies were arrangedsuch that new elements were added to the onesarranged in the previous strategy (e.g., Strategy 3incorporated elements of Strategies 1 and 2 plusadditional elements). Table 2 contains a summa-ry of the elements in each teaching strategy.

Strategy 1. The role of the teacher in Strategy 1was to describe the target relations to the child,arrange the environment to promote activeengagement, and provide feedback when thechild responded correctly. The child and teachersat on the floor across from each other with thetoys on a colored mat that corresponded toStrategy 1 between them. Because this was theonly strategy in which the teacher never prompt-ed responses or provided correct-answer models,presession exposure was arranged in which theteacher vocally labeled each target item once priorto the start of each session. Specifically, theteacher held up one target stimulus at a time andlabeled it in Spanish until each name relation waslabeled once. All interactions were child initiated;the teacher provided no prompts to play and didnot directly or indirectly question the child about


either color- or animal-name relations (i.e., whenthe child touched a target item, thereby initiatinga learning opportunity, the teacher did not deliverprompts of any sort). If the child emitted thetarget response within 5 s of the initiation of thelearning opportunity (e.g., the child said, ‘‘This isazul’’), the teacher provided praise (e.g., ‘‘That’sright, that is azul!’’). In addition, the teacher neverdelivered a model prompt following an error(e.g., the child said, ‘‘This is rosa,’’ while holdinga blue crayon).

Strategy 2. Two additions to the proceduresoutlined for Strategy 1 were included inStrategy 2. First, the teacher provided vocalprompts to name colors and animals when alearning opportunity was initiated (i.e., whenthe child touched a target item for the first timewithin a 15-s interval). Second, the teacherprovided a model of the correct responsefollowing an error and provided an opportunityfor the child to echo the model. When the childinitiated a learning opportunity by touching atarget item such as a red car, in English theteacher asked, ‘‘What color is that car?’’ If thechild said, ‘‘rojo,’’ the teacher provided praise; ifthe child said anything else or did not respond,the teacher provided a model prompt, ‘‘It’srojo.’’ There was no explicit instruction toimitate the teacher’s model prompt, but whenthe child did correctly imitate the model, theteacher provided praise. When the child did notcorrectly imitate the model, the teacher simplycontinued playing. If the child touched a secondtarget item in the same 15-s interval, another

question was not issued; only the first targetitem touched in each 15-s interval occasionedteaching. Two changes from Strategy 1 werealso relevant. First, the relations were notdictated to the child at the start of thesesessions, and second, a different-colored matwas present during Strategy 2.

Strategy 3. The child and teacher sat on thefloor across from each other with a colored matcorrelated with Strategy 3 between them, and abox (25 cm by 40 cm by 50 cm) that containedrelevant toys was next to them. This strategyinvolved two distinct components; the first wasconsistent with direct instruction (teacher-initiat-ed trials, prompt-delay procedures, and differen-tial reinforcement), and the second was consistentwith the teaching procedures described forStrategy 2. Teacher-initiated learning opportuni-ties were conducted in the first component. Theteacher first implemented a 0-s prompt delay,such that a model of the correct response (e.g.,‘‘blanco’’) immediately followed the initialprompt (i.e., ‘‘What color is this?’’). Theprompt-delay schedule then progressed by 1 sin each subsequent session (independent ofresponding) until a 5-s delay was reached. Threesessions were then conducted at the 5-s delay. Ifthe child did not reach the mastery criterionduring these sessions, the delay was reset to 0 sand progressed on a slower schedule (each delaywas implemented for two sessions instead of one).

During each teacher-initiated learning oppor-tunity, the teacher held up a color or animal cardand in English asked, ‘‘What color is this?’’

Table 2

Summary of the Elements of Each Teaching Strategy

Strategy 1 Strategy 2 Strategy 3

Presession exposure — —— — Teacher-initiated learning opportunities

(prompt delay, conditioned and back-upreinforcers)

Child-initiated learning opportunities Child-initiated learning opportunitiesfollowed by a teacher vocal prompt

Child-initiated learning opportunitiesfollowed by a teacher vocal prompt

Praise provided for a correct response Praise provided for a correct response Praise provided for a correct response— Corrective feedback provided for an error Corrective feedback provided for an error

Note. Dash indicates the absence of the element in each teaching strategy.


Following a correct response to the initial vocalprompt, the child received praise and two goldtokens. A model of the correct response and anopportunity to echo the model followed errors.Following a correct response to the modelprompt, the child received praise and one goldtoken. Errors following the model prompt wereignored, and the next trial was initiated. Eachrelation was presented to the child twice; thus,eight teacher-initiated trials were conducted.When the child answered correctly after theinitial vocal prompt of the first presentation of arelation, that color or animal card was notpresented again, such that the total number oflearning trials was reduced. Therefore, thenumber of teacher-initiated learning opportuni-ties ranged from four to eight, depending on thechild’s responding. Once the child received eightgold tokens, he or she was allowed to exchangethem for access to toys in the box (i.e., backupreinforcers) that was decorated and referred to asthe treasure box. The toys in the treasure boxincluded the target stimuli for the same color-and animal-name relations that were targeted inthe first component. This second component ofStrategy 3 was then conducted identically toStrategy 2 for a period of 4 min. The timerequired to complete the first component ofStrategy 3 varied between 30 and 120 s depend-ing on responding. To calculate the total time ofStrategy 3 sessions, the time required to completethe first component was added to the 4 minrequired for the second component. To keep thesession time consistent across strategies, thesession times of Strategies 1 and 2 were yokedto the time required to conduct the previousStrategy 3 session. For example, if it took 1 minand 4 min to conduct the first and secondcomponents of a Strategy 3 session, the followingsessions of Strategies 1 and 2 were both 5 min induration. To establish the yoking procedure, eachassessment started with a Strategy 3 session.

Preference Assessment

As noted above, three distinct sessions madeup a session block, and one session block was

conducted per day. Session blocks alternatedbetween forced choice and free choice. Duringforced-choice blocks, the experimenter random-ly determined the order of the teachingstrategies; during free-choice blocks, the childdetermined the order. The free-choice blocksyielded our measure of children’s preference forthe teaching strategies.

On the outside of the session room door,there were three colored cards, each of whichcorresponded to one of the teaching strategies.The cards were attached to the door in a row,and their placement was randomized eachsession. When the child removed one of thecolored cards from the door and handed it tothe teacher (initial link of the concurrent-chainsarrangement), he or she entered the room toexperience the correlated teaching strategy(terminal link of the chain). At sessioncompletion, the teacher informed the child thatthe session was over and instructed the child tostand up. The teacher and child left the roomfor approximately 30 to 60 s. During this timethe teacher and child played in the hallway (e.g.,passed a ball to each other, or the child mayhave told the teacher a story). These procedureswere repeated until the child experienced eachof the three teaching strategies, and the sessionblock was complete. During the forced-choiceblocks, the teacher stood behind or next to thechild and said, ‘‘Hand me the [color] card.’’These session blocks were arranged to teach thechildren the association between selecting aparticular card and experiencing the correlatedteaching context and to provide evidence of therelative efficacy of the procedures. The maindifference between the forced- and free-choiceblocks was that the teacher said, ‘‘Hand me thecard that you would like to do first [next]’’during the free-choice blocks. All of thechildren followed the instruction to removeand hand a card to the teacher during all sessionblocks. The teacher delivered a short statementof praise following all card selections (i.e., no


differential consequences for selecting a partic-ular card were provided other than access to thedifferent teaching strategies). In both theforced- and free-choice blocks, the selected cardwas removed from the array such that fewercards were present during subsequent selectionopportunities in each session block.

Each assessment continued until the childreached a mastery criterion in one of theteaching strategies or 90 sessions occurred.The mastery criterion was reached when thechild was 100% correct with respect to eachtarget relation for two nonconsecutive sessionsor 80% correct with respect to each targetrelation for three nonconsecutive sessions.

Posttests

Four posttests comprised of 48 trials wereconducted with the color- and animal-namerelations with all children following completionof each assessment. The teacher who conductedthe efficacy and preference assessments alsoconducted two of the posttests. The teacher saton the floor across from the child and held upindividual target objects used previously (e.g.,cars, crayons) and asked, ‘‘What color is this?’’The teacher who conducted the pretestsconducted the other two posttests. Sitting acrossfrom the child at a table, the teacher held upone laminated color or animal card at a timeand said, ‘‘What color is this?’’ Both teachersinstructed the children to answer in Spanish,and no consequences were delivered followingcorrect or incorrect responses. The teachersprovided descriptive praise (e.g., ‘‘I like yourshirt today’’) following every other trial. Correctresponses were tallied, and the results of all fourposttests were added together and divided byfour to obtain a mean posttest score.

RESULTS

The patterns observed in Emma’s (Figure 1),Jeff’s (Figure 2), and Lisa’s (Figure 3) session-by-session data are most representative of thepatterns observed with the other children;

therefore, only these session-by-session dataare depicted. The other children’s data aresummarized in Figure 4 and Table 3.

Efficacy

Figures 1 to 3 depict individual perfor-mance during the color-name assessments inthe first columns and the animal-nameassessments in the second columns. Emma’sresponding (Figure 1) was consistent acrossboth taught relations, with the highest numberof learning opportunities and correct responsesobserved in Strategy 3. She did not reach themastery criterion in either assessment; there-fore, both assessments were terminated fol-lowing the completion of 90 sessions. Jeff(Figure 2) experienced more learning oppor-tunities and emitted the highest number ofcorrect responses in Strategy 3 during bothname-relation assessments. He emitted morecorrect responses in Strategy 2 sessions duringthe animal-name relation assessment than inStrategy 2 sessions in the color-name relationassessment. The asterisks above the data pointsdenote the sessions in which he met masterycriterion in Strategy 3 during both assess-ments. For Lisa (Figure 3), consistent with theother children’s data, the highest number oflearning opportunities was observed in Strat-egy 3 during both assessments. In addition,she emitted more correct responses in Strategy3 than in Strategies 1 and 2 across both name-relation assessments. The mastery criterion wasmet in Strategy 3 during the color-namerelation assessment; however, the masterycriterion was not met during the animal-namerelation assessment.

Group means (and standard deviations) ofthe efficacy measures are shown in Figure 4.The mean number of learning opportunitieswas roughly equivalent between Strategies 1 and2 (Ms 5 8.3 and 8.6 learning opportunities persession, respectively). By contrast, a highermean number of learning opportunities wasobserved in Strategy 3 (M 5 13.4). Childrenrarely emitted correct responses in the absence


of teacher vocal prompts, as evident by the lowmean number correct in Strategy 1 (M 5 0.3).Although the mean number correct was higherin Strategy 2 (M 5 3.1) than in Strategy 1, themean number correct observed in Strategy 2was less than half that of Strategy 3 (M 5 6.8).Table 3 shows that the highest percentage ofcorrect responding was observed in Strategy 3

for all children. In addition, the masterycriterion was met in 7 of the 12 name-relationassessments exclusively in Strategy 3.

In addition to analyzing acquisition data, wealso inspected pre- and posttest scores asadditional indicators of the relative efficacy ofthe teaching strategies. A small, statisticallyinsignificant difference in the mean percentage

Figure 1. The cumulative number of learning opportunities (top panels), cumulative number of correct responses(second panels), preference rank (third panels), and pre- and posttest scores (bottom panels) during the efficacy andpreference evaluations for Emma.


correct was found across the two differentteachers who conducted the posttests; thus thescores from the four posttests were combinedand reported as a mean. Figure 1 (bottom)shows Emma’s mean pre- and posttest scores.She scored 0% correct on pretests with respectto all relations. Posttest scores increased with

respect to all relations, but the highest posttestscores observed were for relations taught inStrategy 3. Jeff’s mean pre- and posttest scoresare presented in Figure 2 (bottom). He alsoscored 0% correct with respect to both sets ofrelations. His mean posttest scores for the color-name relations show an increase in percentage

Figure 2. The cumulative number of learning opportunities (top panels), cumulative number of correct responses(second panels), preference rank (third panels), and pre- and posttest scores (bottom panels) during the efficacy andpreference evaluations for Jeff.


correct with respect to Strategy 1 and 3 relations.His mean posttest scores for the animal namesincreased with respect to all relations. Althoughhe met the mastery criterion in Strategy 3 duringthe efficacy assessment, the highest posttestscores were observed for the relations taught inStrategy 2. Lisa’s posttests for the color-namerelations were not conducted in the same manner

in which the other children’s posttests wereconducted due to teacher error; thus, they are notincluded in the current analysis. However, Lisa’smean pre- and posttest scores for the animal-name relations are shown in Figure 3 (bottom).She scored 0% correct with respect to all taughtrelations, and her mean posttest scores increasedfrom pretest, with the highest percentage correct

Figure 3. The cumulative number of learning opportunities (top panels), cumulative number of correct responses(second panels), preference rank (third panels), and pre- and posttest scores (bottom panels) during the efficacy andpreference evaluations for Lisa.


observed with respect to the relations taught inStrategy 3.

Figure 4 shows the mean posttest scoresacross all children. (All children’s pretest scoreswere zero; thus all posttest scores represent bothan absolute score as well as a percentage changescore.) The overall mean number correct duringthe posttests was higher with respect to Strategy1 relations than Strategy 2 relations (Ms 5 7.2

and 5.7, respectively); however, the highestmean number correct was observed with respectto Strategy 3 relations (M 511.9).

The mean number of errors, which weredefined as learning opportunities withoutcorrect responses, occurred more often inStrategy 3 (M 5 6.6, SD 5 1.7) than inStrategy 2 (M 5 5.1, SD 5 1.2). It is importantto point out that in addition to more errors,

Figure 4. Mean number of learning opportunities, correct responses during acquisition, correct responses onposttests, and mean preference rank across all teaching strategies. The lines above each bar represent thestandard deviations.

Table 3

Efficacy and Preference Assessment Results Summary

Child RelationHighest percentage

correct Highest posttest score Most preferred Least preferred

Emma Color 3 3 3 1Animal 3 3 3 1

Mary Color 3 1 3 2Animal a 3 2/3 3 2

Quinn Color 3 3 2 3Animal a 3 3 3 2

Jeff Color a 3 3 1 2Animal a 3 3 3 2

Rena Color a 3 3 1 3Animal a 3 3 3 2

Lisa Color a 3 N/Ab 1 2Animal 3 3 1 3

Mode outcomes 3 3 3 2

a Indicates the assessment in which the child reached the mastery criterion within 30 instructional sessions.b Indicates the exclusion of the posttest results due to procedural inconsistencies.


there were more learning opportunities inStrategy 3; thus, the mean proportion of errors,derived by dividing the number of errors by thenumber of learning opportunities, was actuallyhigher in Strategy 2 (M 5 .66) than in Strategy3 (M 5 .46).

The mean amount of time each childexperienced each strategy was 112 min per setof relations (range, 37 min to 155 min). InStrategy 3, the mean amount of time in the firstcomponent was only 22 min (range, 9 min to35 min), and the mean duration of the secondcomponent was 90 min (range, 28 min to120 min).

Preference

Emma’s preference rank of the teachingstrategies across free-choice opportunities isdisplayed in the third pair of panels in Figure 1.A rank of 1 represents the teaching strategy thatwas selected first, whereas a rank of 3 representsthe strategy that was selected last during eachfree-choice block. Selections during the color-name relation assessment were initially variable;however, she consistently selected Strategy 3first during the last 11 free-choice blocks. Sheselected Strategy 3 first on 13 of 15 opportu-nities, although some variability in selectionswas evident during the animal-name relationassessment. Jeff’s preference assessment data aredepicted in the third pair of panels in Figure 2.He selected Strategy 1 first during five of theeight free-choice blocks during the color-namerelation assessment. However, during the ani-mal-name relation assessment, he selectedStrategy 3 first during 8 of the 10 free-choiceblocks. Figure 3 shows Lisa’s preference assess-ment results in the third pair of panels. Duringboth assessments, she selected Strategy 1 firstalmost exclusively during all free-choice blocks.

Table 3 shows that 2 of the 6 children(Emma and Mary) showed a relative preferencefor Strategy 3 during both color- and animal-name assessments. Three of the 6 children(Quinn, Jeff, and Rena) initially showed arelative preference for either Strategy 1 or 2

during the first assessment (color); however, all3 children showed a relative preference forStrategy 3 during the second assessment(animal). Lisa showed a relative preference forStrategy 1 during both assessments. In sum,following experiences with each teaching strat-egy during the color-name relation assessment,5 of the 6 children showed a preference forStrategy 3 during the animal-name relationassessment, and Strategy 2 was least preferredfor 4 of the 6 children following this sameexperience.

DISCUSSION

We determined that the mixed approach(Strategy 3) involving discovery, embedded, anddirect instruction was the most efficacious forteaching preschool children name relations. Byarranging teacher-initiated learning opportuni-ties, a mean of 115 more learning opportunitieswere experienced per relation during Strategy 3than in Strategies 1 and 2. As a result, Strategy 3produced the highest number of correctresponses, the least amount of time to reachthe mastery criterion, and the highest posttestscores.

In the absence of explicit teacher prompts,the children rarely emitted correct responses inStrategy 1 sessions. Nevertheless, childrenscored a mean of 41% correct with respect toStrategy 1 relations on our posttest measures.Furthermore, the results of 5 of the 11 posttestsindicated that children scored higher withrespect to Strategy 1 relations than Strategy 2relations. These results suggest that simplyexposing children of typical development totarget relations in the absence of requirementsfor responding was sufficient for some learningto occur.

Mastery was, however, met only in Strategy3. This occurred in 7 of 12 of the applications.Thus, after a mean of only 22 min of directinstruction was provided in addition to discov-ery-oriented and embedded teaching strategies,there was more than a 50% chance that a


concept class would be mastered. By contrast,the probability of mastering a concept classwhen only discovery-oriented or embeddedteaching strategies were implemented for asimilar amount of time was zero. By arrangingfor some intermittent and brief teacher-initiatedlearning opportunities, measurable gains inlearning were achieved.

The procedures implemented in Strategy 3were perhaps more effective in promoting skillacquisition due to the interaction between theprompt-delay procedures and the motivationalsystem implemented during the teacher-initiat-ed component. Implementation of the prompt-delay procedure most likely allowed transfer ofstimulus control from the model prompt to theactual stimuli (Wolery & Gast, 1984). Thecontingent delivery of tokens exchangeable fortoys (which occasioned child-initiated learningopportunities) probably provided sufficientmotivating conditions for learning to occur aswell. It is also possible that the relative efficacyof Strategy 3, especially with regard to theposttest outcomes that could be construed asgeneralization measures, was predicated on theinteraction between the multiple teachingstrategies implemented during Strategy 3.

Both teacher-initiated and child-initiatedlearning opportunities were included in Strategy3 to facilitate both the acquisition and general-ization of target relations. Some early childhoodresearchers have expressed a concern that embed-ded teaching procedures may not in isolationprovide sufficient learning opportunities for skillacquisition (e.g., Daugherty et al., 2001; Van DerHeyden, Snyder, Smith, Sevin, & Longwell,2005). Our results provide support for thoseconcerns. Presumably due to children’s limitedinteraction with specific toys or in specificcontexts (e.g., Cammilleri & Hanley, 2005),only some target relations may be taught duringdiscovery-oriented and embedded teaching strat-egies. By analyzing the percentage of sessions inwhich all four target relations were occasioned atleast once within each teaching strategy, we found

that all four relations were occasioned an averageof 47% (range, 10% to 81%), 26% (range, 0% to47%), and 100% of Strategy 1, 2, and 3 sessions,respectively. Thus, by arranging for short periodsof direct instruction, all of the target relationswere represented in all teaching sessions, therebyensuring that learning opportunities for all targetrelations would be experienced.

Perhaps more surprising than the efficacyoutcomes were the outcomes with respect tochildren’s preferences for the different teachingcontexts. When approximately 20% of teachingtime was usurped by teacher-directed instruc-tion in Strategy 3, children did not avoid thisstrategy. Instead, 5 of 6 children showed apreference for this strategy when the secondrelation was taught. (Note that although wemeasured children’s preference for the strategieswhile teaching both relations, we find thepreference data with respect to the secondrelation taught more compelling, given that thechildren had more experience with each strategyat these points in time.) It is also important topoint out that our data showed that embeddedteaching (Strategy 2), which is recommended asbest practice for teaching young children(Bredekamp & Copple, 1997; Bricker et al.,1998), was the least preferred for 4 of the 6 ofthe children during the second assessment.

Lisa preferred Strategy 1 during both the firstand second assessments. Her data are unique inthat she selected Strategy 1 first in her initialfree-choice block, and her selections nevervaried. By contrast, preferences emerged overtime for the other 5 children. Lisa’s data suggestthat either she did indeed have a strongpreference for discovery-oriented teaching fol-lowing a single experience with that strategy, orsome other variable controlled her selectionsabove and beyond the programmed conse-quences for card selections (e.g., color bias,self-generated rule).

It was our goal to arrange ecologically validteaching conditions that varied in teacherdirectedness; because of this, the three options


differed in multiple ways. Understanding thecontrolling variables for the observed preferenc-es is complicated by this fact. Furthermore, anyindividual child’s preference may have been adynamic interaction between, for example, thepotentially reinforcing elements of Strategy 3and the potentially aversive elements of Strategy2. The reinforcing elements of Strategy 3 mayhave been the inclusion of conditioned andbackup reinforcers, the relatively high numberof descriptive praise statements that were afunction of the higher amount of correctresponding, the varied nature of the teaching(i.e., all three strategies were experienced), orthe fact that children were simply more effectiveunder these conditions. The identified elementsmay have been operating independently or incombination to influence children’s preferencefor Strategy 3.

Although there were more total errors inStrategy 3 than in Strategy 2, there were alsomany more learning opportunities and morecorrect responses in Strategy 3. Thus, there werea higher proportion of errors in Strategy 2, andit is possible that the high proportion of errorsexperienced in Strategy 2 may have led childrento avoid this teaching context. Our data areconsistent with other learning research that hasdemonstrated the aversive properties of condi-tions associated with high levels of errors. Forinstance, while examining the effects of taskdifficulty on the aberrant behavior of 2 childrenwith severe developmental disabilities, Weeksand Gaylord-Ross (1981) found that tasks thatresulted in more errors were also associated withhigher levels of aberrant behavior to escape thetask. Because of the possible impact of thisvariable on children’s preferences for instruc-tional strategies, future research should examinethe effect the number and proportion of errorshas on children’s preferences for teachingconditions in a more controlled manner.

Another potentially aversive element ofStrategy 2 may have been the delivery ofinstructions while children interacted with

preferred activities. It is possible, and we thinklikely, that the teacher prompts in Strategy 2represented a brief time-out from preferredactivities in that we were repeatedly interruptingchildren’s play to deliver instructions during ourembedded teaching. We did indeed observe that1 child (Quinn), who selected Strategy 2 theleast during his second preference assessment,touched the target toys less across time duringStrategy 2 sessions. Because no teacher ques-tions were provided for touching toys inStrategy 1 and touching the target toys persistedduring Strategy 1 sessions, preliminary evidencethat the embedded prompting in Strategy 2 wasaversive is apparent.

The somewhat speculative nature of ourassertions regarding controlling variables forpreference, especially as they relate to embeddedteaching, has occasioned a number of researchquestions. We think it is critical to first evaluatedifferent embedded teaching strategies that varyin the rate and proportion of prompts anddescriptive comments regarding play. In addi-tion, it is likely that the preference value of thetoys and the initial skill difficulty are influential.For example, attempting to teach a highlydifficult skill to a child while she is playing withher most preferred toys is likely to create anonpreferred teaching context. Therefore, theinfluence of these factors on efficacy andpreference should be evaluated.

The children in the current investigationwere all of typical development; however, webelieve that our results have implications forchildren at risk or with developmental disabil-ities. Although individual studies have demon-strated the efficacy of direct instruction (e.g.,Risley & Wolf, 1967), more recent studies haveisolated relative advantages of normalizedapproaches such as incidental teaching (e.g.,McGee et al., 1985), natural language teachingparadigm (e.g., Koegel, Koegel, & Surratt,1992), or pivotal response training (e.g.,Schreibman, Kaneko, & Koegel, 1991) forteaching language to young children with


autism or related disabilities. Our results suggestthat direct teaching clearly adds value to thenormalized approaches, and that language goalsmay be difficult or impossible to achievewithout some direct teaching. Therefore, ourresults support the general assertion thatcombining the two language interventionprocedures will yield optimal results for lan-guage learners (e.g., Lennox & Brune, 1993;Schepis et al., 1982).

We identified children’s preferences forteaching contexts using procedures similar tothose used to identify children’s preferences forbehavioral interventions (e.g., Hanley et al.,1997, 2005). Our methods differed fromprevious applications of concurrent-chains ar-rangements in multiple ways. In previousstudies, experiences within each relevant contextfor which preference was to be assessed wereconducted prior to preference assessments toensure that discriminated selections wouldemerge when the opportunity to choose wasarranged (Hanley et al., 1997, 2005). Bycontrast, experience with the contexts andassessment of preference occurred simultaneous-ly and from the beginning of each evaluation inthe current study. Thus, we were able to assesspreferences early and as they emerged as afunction of experiencing the teaching strategies.Our study also differed by the manner in whichthe terminal links were made available followingselections. That is, following a selection of aparticular strategy, that strategy was restrictedfrom the subsequent choice opportunity (bycontrast, the stimulus array in the initial linksremained constant in Hanley et al., 1997,2005). Although a child’s motivation to selecthis or her most preferred context could beaffected by the fact that he or she will alwaysexperience each context for the same amount oftime, this restriction procedure allowed apreference hierarchy to develop in addition toidentifying the most preferred teaching strategy,and it also ensured that children were exposed

to each teaching strategy for the same amountof time. These procedural modifications result-ed in a more comprehensive description ofpreferences and ensured the experimentalintegrity of the comparative efficacy analysis.Thus, our assessment allowed us to determinedirectly the acceptability of classroom practiceswith the children who experience the practices.It is our hope that early childhood researcherswho conduct comparative analyses of practicesor interventions will consider determiningchildren’s preferences in a similar manner inaddition to determining relative efficacy in theirevaluations.

The selection of teaching strategies for youngchildren has been, and to some extent continuesto be, based on developmentally appropriatepractice (Bredekamp, 1987), with the promi-nent strategies being of child-initiated orienta-tion almost to the exclusion of more teacher-initiated direct-instruction strategies (Carta,Atwater, Schwartz, & McConnell, 1993; Carta,Schwartz, Atwater, & McConnell, 1991; John-son & Johnson, 1992). More recent efforts havebeen made to identify a range of evidence-basedteaching strategies (Smith et al., 2002). Aprimary contribution of the current study is topoint out that the efficacy of each strategy isonly one important measure; child preferencefor strategies under consideration should also betaken into account. Although recommendationsto assess the social validity of interventions havebeen made (Schwartz, 1999; Schwartz & Baer,1991; Wolf, 1978), Odom and Strain (2002)reported that of the 184 single-subject studiesidentified in the child-focused recommendedpractices strand of the DEC task force, only15% and 27% of studies assessed the socialvalidity of the procedures and outcomes,respectively, and to our knowledge noneassessed social validity directly (i.e., all reliedon verbal reports) or with the childrenthemselves. Thus, we are calling for more andbetter assessments of social validity to beconsidered during the evolution of earlychildhood recommended practices.


Because we evaluated general strategies asopposed to specific tactics (e.g., a brief time-outvs. a correct model following incorrect respons-es), the manner in which we designed the threeteaching strategies may certainly differ to somedegree from the manner in which thesestrategies are implemented in many preschoolclassrooms. A good direction for future researchwould be to incorporate different types of theseteaching strategies into comparative analysessuch that the conditions under which eachstrategy may confer advantages can be articu-lated. For instance, various forms of discovery-oriented teaching often occur in preschoolclassrooms out of necessity, given that teacher-to-child ratios vary from 1:3 to 1:10, dependingon the age of the children. The efficacy of thistype of teaching undoubtedly varies as afunction of the likelihood of the materialsselected to evoke developmentally importantbehavior and provide automatic reinforcementfor those behaviors (for good examples, seematerials originally described by Maria Mon-tessori; Martin, 1993). Furthermore, our Strat-egy 2 involved a specific amount and timing ofteacher questions and model prompts in thecontext of a particular play context. All of theseparameters should be varied and evaluated infuture research.

It is also important to note that we evaluatedthe relative efficacy of Strategy 2 in the context ofa one-on-one teaching situation as opposed to thecontext of ongoing classroom activities with otherchildren and teachers present, which is typicallythe context in which embedded teaching occurs.However, Pretti-Frontczak and Bricker (2001)found that when teachers embedded learningopportunities (which was not often), they did soduring one-on-one activities with the child. Thus,we do not believe that our controlled arrange-ment greatly detracts from the generality of ourresults, and we contend that variations of theseteaching strategies should be evaluated undersimilarly controlled, and thus technologicallydescribable, conditions in future research.

Our research goal was to identify empirically aset of teaching procedures that was most effectiveand preferred by the children who experiencedthe strategies. Our results, although preliminary,support the use of varied teaching practices thatincorporate discovery-oriented, embedded, anddirect-instruction approaches. Thus, we recom-mend that early childhood and early childhoodspecial education teachers arrange some teacher-initiated learning opportunities throughout theday, in addition to exclusively or partially child-initiated learning opportunities, to promote skillacquisition and to provide learning environmentsthat are preferred by children.

REFERENCES

Adams, G. L., & Englemann, S. (1996). Research on directinstruction: Twenty-five years beyond Distar. Seattle:Educational Achievement Systems.

Bredekamp, S. (1987). Developmentally appropriate practicein early childhood programs serving children from birththrough age 8. Washington, DC: National Associationfor the Education of Young Children.

Bredekamp, S., & Copple, C. (1997). Developmentallyappropriate practices in programs for children 3 to 8.Washington, DC: National Association for theEducation of Young Children.

Bricker, D., Pretti-Frontczak, K., & McComas, N. R.(1998). An activity-based approach to early intervention(2nd ed.). Baltimore: Brookes.

Bruner, J. S. (1961). The act of discovery. HarvardEducational Review, 31, 21–32.

Cammilleri, A. P., & Hanley, G. P. (2005). Use of a lag-differential reinforcement contingency to increasevaried selections of classroom activities. Journal ofApplied Behavior Analysis, 38, 111–115.

Carta, J. J., Atwater, J. B., Schwartz, I. S., & McConnell,S. R. (1993). Developmentally appropriate practice: Areaction to Johnson and McChesney Johnson. Topicsin Early Childhood Special Education, 13, 243–254.

Carta, J. J., Schwartz, I. S., Atwater, J. B., & McConnell,S. R. (1991). Developmentally appropriate practice:Appraising its usefulness for young children withdisabilities. Topics in Early Childhood Special Educa-tion, 11, 1–20.

Catania, C. A., & Sagvolden, T. (1980). Preference forfree choice over forced choice in pigeons. Journal ofthe Experimental Analysis of Behavior, 34, 77–86.

Daugherty, S., Grisham-Brown, J., & Hemmeter, M. L.(2001). The effects of embedded skill instruction onthe acquisition of target and nontarget skills inpreschoolers with developmental delays. Topics inEarly Childhood Special Education, 21, 213–221.


Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L.P., Owens, J. C., & Slevin, I. (1992). A comparisonof two approaches for identifying reinforcers forpersons with severe and profound disabilities. Journalof Applied Behavior Analysis, 25, 491–498.

Fox, L., & Hanline, M. F. (1993). A preliminaryevaluation of learning within developmentally appro-priate early childhood settings. Topics in EarlyChildhood Special Education, 13, 308–327.

Fredrick, L. D., Deitz, S. M., Bryceland, J. A., &Hummel, J. H. (2001). Behavior analysis, education,and effective schooling. Reno, NV: Context Press.

Hanley, G. P., Piazza, C. C., Fisher, W. W., Contrucci, S.A., & Maglieri, K. A. (1997). Evaluation of clientpreference for function-based treatment packages.Journal of Applied Behavior Analysis, 30, 459–473.

Hanley, G. P., Piazza, C. C., Fisher, W. W., & Maglieri,K. A. (2005). On the effectiveness of and preferencefor punishment and extinction components offunction-based interventions. Journal of AppliedBehavior Analysis, 38, 51–65.

Hart, B. M., & Risley, T. R. (1968). Establishing use ofdescriptive adjectives in the spontaneous speech ofdisadvantaged preschool children. Journal of AppliedBehavior Analysis, 1, 109–120.

Hart, B. M., & Risley, T. R. (1975). Incidental teachingof language in the preschool. Journal of AppliedBehavior Analysis, 8, 411–420.

Heal, N. A., & Hanley, G. P. (2007). Evaluatingpreschool children’s preferences for motivationalsystems during instruction. Journal of Applied Behav-ior Analysis, 40, 249–262.

Horn, E., Lieber, J., Li, S., Sandall, S., & Schwartz, I. S.(2000). Supporting young children’s IEP goals ininclusive settings through embedded learning oppor-tunities. Topics in Early Childhood Special Education,20, 208–223.

Johnson, J. E., & Johnson, K. M. (1992). Clarifying thedevelopmental perspective in response to Carta,Schwartz, Atwater, and McConnell. Topics in EarlyChildhood Special Education, 12, 438–457.

Klahr, D., & Nigam, M. (2004). The equivalence oflearning paths in early science instruction: Effects ofdirect instruction and discovery learning. PsychologicalScience, 15, 661–667.

Koegel, R. L., Koegel, L. K., & Surratt, A. (1992).Language intervention and disruptive behavior inpreschool children with autism. Journal of Autism andDevelopmental Disorders, 22, 141–153.

Lennox, D. B., & Brune, P. (1993). Incidental teachingfor training communication in individuals withtraumatic brain injury. Brain Injury, 7, 449–454.

Losardo, A., & Bricker, D. (1994). Activity-based interven-tion and direct instruction: A comparison study.American Journal on Mental Retardation, 98, 744–765.

Magliaro, S. G., Lockee, B. B., & Burton, J. K. (2005).Direct instruction revisited: A key model forinstructional technology. Educational Technology Re-search and Development, 53, 41–55.

Martin, K. (1993). Preparing for life: Montessori’sphilosophy of sensory education. Montessori Life, 5,24–27.

Mayer, R. E. (2004). Should there be a three-strikes ruleagainst pure discovery learning? American Psychologist,59, 14–19.

McGee, G. G., Krantz, P. J., & McClannahan, L. E.(1985). The facilitative effects of incidental teachingon preposition use by autistic children. Journal ofApplied Behavior Analysis, 18, 17–31.

Odom, S. L., & Strain, P. S. (2002). Evidence-basedpractice in early intervention/early childhood specialeducation: Single-subject design research. Journal ofEarly Intervention, 25, 151–160.

Piaget, J. (1970). Science of education and psychology of thechild. New York: Oxford University Press.

Pretti-Frontczak, K., & Bricker, D. (2001). Use of theembedding strategy during daily activities by earlychildhood education and early childhood specialeducation teachers. Infant Toddler Intervention, 11,111–128.

Risley, T. R., & Wolf, M. M. (1967). Establishingfunctional speech in echolalic children. BehaviourResearch and Therapy, 5, 73–88.

Schepis, M. M., Reid, D. H., Fitzgerald, J. R., Faw, G. D.,Van Den Pol, R. A., & Welty, P. A. (1982). Aprogram for increasing manual signing by autistic andprofoundly retarded youth within the daily environ-ment. Journal of Applied Behavior Analysis, 15,363–379.

Schreibman, L., Kaneko, W. M., & Koegel, R. L. (1991).Positive affect of parents of autistic children: Acomparison across two teaching techniques. BehaviorTherapy, 22, 479–490.

Schwartz, I. S. (1999). Controversy or lack of consen-sus? Another way to examine treatment alternatives.Topics in Early Childhood Special Education, 19,189–198.

Schwartz, I. S., & Baer, D. M. (1991). Social validityassessment: Is current practice state of the art? Journalof Applied Behavior Analysis, 24, 189–204.

Sidman, M. (1960). Tactics of scientific research. Oxford,UK: Basic Books.

Smith, B. J., Strain, P. S., Snyder, P., Sandall, S. R.,McLean, M. E., Ramsey, A. B., et al. (2002). DECrecommended practices: A review of 9 years of EI/ECSEresearch literature. Longmont, CO: Sopris West.

Solter, A., & Mayer, R. E. (1978). Broader transferproduced by guided discovery of number conceptswith preschool children. Journal of EducationalPsychology, 70, 363–371.

Stebbins, L. B., St. Pierre, R. G., Proper, E. C., Anderson,R. B., & Cerva, H. (1977). Education as experimen-tation: A planned variation model. An evaluation ofProject Follow Through. Cambridge, MA: Abt.

Strain, P. S., McConnell, S. R., Carta, J., Fowler, S. A.,Neisworth, J. T., & Wolery, M. (1992). Behaviorismin early intervention. Topics in Early Childhood SpecialEducation, 12, 121–141.


Tiger, J. H., Hanley, G. P., & Heal, N. A. (2006). Theeffectiveness of and preschoolers’ preference forvariations of multiple-schedule arrangements. Journalof Applied Behavior Analysis, 39, 475–488.

Tiger, J. H., Hanley, G. P., & Hernandez, E. (2006). Anevaluation of the value of choice with preschoolchildren. Journal of Applied Behavior Analysis, 39, 1–16.

Van Der Heyden, A. M., Snyder, P., Smith, A., Sevin, B.,& Longwell, J. (2005). Effects of complete learningtrials on child engagement. Topics in Early ChildhoodSpecial Education, 25, 81–94.

Walsh, B. A., & Petty, K. (2006). Frequency of six earlychildhood education approaches: A 10-year contentanalysis of early childhood education journals. EarlyChildhood Education Journal, 34, 301–305.

Weeks, M., & Gaylord-Ross, R. (1981). Task difficulty andaberrant behavior in severely handicapped students.Journal of Applied Behavior Analysis, 14, 449–463.

Wolery, M., & Gast, D. L. (1984). Effective and efficientprocedures for the transfer of stimulus control. Topicsin Early Childhood Special Education, 4, 55–77.

Wolery, M., & Sainato, D. M. (1996). General curriculumand intervention strategies. In S. L. Odom & M. E.McLean (Eds.), Early intervention recommended practices(pp. 125–158). Austin, TX: Pro-Ed.

Wolery, M., & Wilbers, J. (Eds.). (1994). Including child-ren with special needs in early childhood programs.Washington, DC: National Association for theEducation of Young Children.

Wolf, M. M. (1978). Social validity: The case forsubjective measurement, or how applied behavioranalysis is finding its heart. Journal of Applied BehaviorAnalysis, 11, 203–214.

Woods, J., Kashinath, S., & Goldstein, H. (2004). Effectsof embedding caregiver implemented teaching strat-egies in daily routines on children’s communicationoutcomes. Journal of Early Intervention, 26, 175–193.

Received May 21, 2007Final acceptance September 24, 2007Action Editor, Chris Ninness


Documents

AN EVALUATION OF THE RELATIVE EFFICACY OF … manner in which teachers mediate children’s learning varies across ... direct instruction, discovery learning, ... its reliance on contrived